Application of a Second-order Stochastic Optimization Algorithm for Fitting Stochastic Epidemiological Models

TL;DR

This paper introduces PSPO, an advanced stochastic optimization algorithm that leverages parallel computing to efficiently fit complex stochastic epidemiological models, demonstrated on a historical measles outbreak dataset.

Contribution

The paper presents PSPO, an extension of SPSA, optimized for parallel environments, improving likelihood maximization in stochastic epidemiological modeling.

Findings

PSPO significantly outperforms gradient ascent.

PSPO outperforms SPSA in likelihood maximization.

Efficient fitting of complex models using parallel computing.

Abstract

Epidemiological models have tremendous potential to forecast disease burden and quantify the impact of interventions. Detailed models are increasingly popular, however these models tend to be stochastic and very costly to evaluate. Fortunately, readily available high-performance cloud computing now means that these models can be evaluated many times in parallel. Here, we briefly describe PSPO, an extension to Spall's second-order stochastic optimization algorithm, Simultaneous Perturbation Stochastic Approximation (SPSA), that takes full advantage of parallel computing environments. The main focus of this work is on the use of PSPO to maximize the pseudo-likelihood of a stochastic epidemiological model to data from a 1861 measles outbreak in Hagelloch, Germany. Results indicate that PSPO far outperforms gradient ascent and SPSA on this challenging likelihood maximization problem.